1. Field of the Invention
This invention is related to pin fin heat sinks of the type that are used to transfer heat away from a component to the surrounding atmosphere. This invention is specifically related to heat sinks that are used to cool integrated circuit components, but the device and method are also applicable to pin fin heat sinks that are used with other electronic or electrical components, such as multi-chip modules, transformers and power supplies.
2. Description of the Prior Art
Currently pin fin heat sinks are produced by extrusion, casting and metal removal (ie: machining) processes. Each of these processes constrains the heat sink designer by limiting the geometry and material combinations that can be achieved within a given volume. Manufacturers such as Wakefield Engineering Inc., Thermalloy Inc. and Aavid Thermal Technologies produce large quantities of these devices that are used primarily in the electronics industry to cool PC board mounted devices such as microprocessors. Without heat sinks to carry heat away, these devices would suffer high rates of failure. With the relentless trend toward higher density electronic assemblies coupled with smaller high power dissipating devices the need for optimum heat sink performance is greater than ever. The heat sink designer seeks to optimize heat transfer by generating designs having the largest hA product within the volume permitted, where h is the convection heat transfer coefficient and A is the wetted surface area.
Pin Fin Surface Area
Commonly the cross sections of extruded and machined pin fin heat sinks are rectangular and have two opposing sides formed from the extrusion process and two additional sides formed by a metal removal process such as gang sawing, milling or punching, as shown in U.S. Pat. No. 5,572,789. If the rectangular cross-section is instead longitudinally fluted, the wetted surface area can be increased. Currently produced cast heat sinks commonly provide smooth cylindrical or conical fins. If cylindrical or conical fins have threads formed on the outer surface, wetted area can be increased and the non-smooth thread surface may promote mixing of the fluid flow and thus improve performance.
Pin Fin Spacing
Fin spacing is the distance between the surfaces of any two adjacent fins. It can also be thought of as the width of air channel between fins. Machined, cast and extruded pin fin heat sinks are limited in their fin spacing to approximately 2 mm. This limitation is imposed by tool breakage problems that arise when narrower gaps are attempted. This limitation imposes an obstacle for the heat sink designer who may wish to reduce the gap depending on air flow expectations. Reduction of the gap permits denser fin spacing thereby increasing overall pin fin heat sink surface area and in some conditions o fair flow increased heat transfer performace. The method described below allows a smaller fin to fin gap than can be achieved in conventional processes. This close spacing of fins may be especially beneficial when fins of non-constant cross section (eg: spherical) are employed. Additionally, fins need not be located in rectangular arrays but can be positioned arbitrarily if the designer so chooses. Fin spacing flexibility allowing narrow channel designs enables the designer more freedom to enhance h, the convection heat transfer coefficient.
Pin Fin Material
Most commonly used in conventional heat sinks are metallic extrusion and casting alloys which are tailored for their particular process but which exhibit less than optimum thermal characteristics. Typically purer metals possess superior thermal conductivity than their alloyed cousins. Using the method described below practically any pure metal or alloy can be used as a fin material. The only criteria required for fin and base material use is weldability. Even dissimilar base and fin materials can be successfully joined. While heat sinks are typically convection limited, a high thermal conductivity of the heat sink structure will beneficially effect overall heat transfer. Conversely, if conductivity of the fin/base joint is lower than parent metal as when the joint is prepared by soldering or brazing or swaging, overall performance may be diminished.
Pin Fin Length
Extruded and cast pin fin heat sinks commonly have pins exhibiting a height to pin spacing ratio of less than 10 to 1, as shown in U.S. Pat. Nos. 4,879,891 and 4,884,331. Process limitations prevent larger ratios. The method described below can achieve pin fin height to spacing ratios that are orders of magnitude greater than those seen in conventional heat sinks. Larger ratios free the heat sink engineer to create more efficient new designs. Longer fins are generally preferred in natural convection circumstances. Fin length may affect both parameters, h and A.
Pin Fin Thickness
Pin fin thicknesses in conventionally manufactured heat sinks range down to approximately 2 mm. In the new method fin thickness can be reduced significantly to approximately 0.25 mm. Greater flexibility in choosing fin thickness provides an advantage in achieving maximum heat transfer.
Pin Fin Orientation
Conventional pin fin heat sink fin axes are parallel, as mandated by the processes employed in their manufacture. The method described below permits non-parallel fin axes and consequently highly variable fin spacings in heat sinks.
Base Thickness
Conventional pin fin heat sink bases are relatively thick compared to fin thickness. The thicknesses that occur are required to some extent by the method of manufacture. However, optimum thermal performance may call for thinner bases to minimize thermal resistance. The method described below permits base thickness to be reduced to approximately 0.25 mm, thus it is possible to weld fins directly onto an exposed copper layer of a PC board itself.
The invention described herein would be suitable for use with pin fins and pin fin configurations that would have superior heat transfer characteristics relative to conventional pin fin heat sinks that are produced by such methods as extrusion, casting or material removal.
The method described below permits the use of cylindrical and spherical fins as well as pin fins having a wide variety of constant (eg: cylindrical, elliptical, fluted) and non-constant cross sections (eg: spherical, knurled, helical,and tapered). For example this invention could be employed with pins having helical grooves or threads extending around the periphery of the pins.
A pin fin heat sink according to this invention includes a base with a plurality of pins extending from the base. Each pin is joined to the base by an autogenous stud weld. The stud weld forms a continuous weld layer formed from material contributed by the pin and material contributed by the base and extends over a complete cross section of the pin to join each pin to the base. The pin is welded to the base by joining molten material from an end of each pin and from the portion of the base adjacent the end of the pin after which the molten material solidifies to form the continuous weld layer.
In the method of fabricating a pin fin heat sink, individual pins are sequentially positioned adjacent a base. A voltage difference is then applied between individual pins and the base sufficient to cause an arc between the pin and the base to melt ends of pins adjacent the base and to melt portions of the base adjacent pins so that the pins are welded to the base.